Automation

For the first couple flights, we'll use roving robots for the long-range exploration and to shelter the habitat after the crew is gone. In fact, we'll probably always using rovers for long-range exploration, at least for several decades.

Automation for fine-manipulation tasks is less cost-effective. When you have to design things to be manipulated by robots, the cost and weight go way up. To get a full appreciation of this, find a robotics lab and try working the waldoes yourself. That's where torque reaction fittings and alignment guides and positioning targets come into play; suddenly a one-ounce bolt weighs five pounds.

Automating functions inside the spacecraft, like throwing switches and turning knobs, can be done better by providing a general purpose telerobot which is able to reach all those switches and knobs. They tried this with the Charlotte robot on the last SPACEHAB flight with great success. One option to consider is having our crew set up a Charlotte just before they button up the module. Alternatively, we could use lightweight manipulators sort of like miniature versions of the Shuttle RMS. (This is said with the space station RMS in mind; but most folks may not be intimately familiar with that design. The primary design difference is staggered joints and redundant motors.)